CONTRIBUTION OF METOP ASCAT DATA FOR LAND SURFACE PARAMETERS MONITORING OVER THE SAHEL Frison P.-L. (1), Faye G. (1)(2), Riazanoff S. (1)(3), Mougin E. (4), Jarlan L. (4), Baup F. (4), Hiernaux P. (4), Ndione J.-A. (5),Wade S. (2), Rudant J.-P. (1), 1. Université Paris-Est Marne-la-Vallée 2. Université Cheikh Anta Diop de Dakar 3. VisioTerra 4. CESBIO 5. Centre de Suivi Ecologique Sénégal (CSE) AMMA - African Monsoon Multidisplinary Analyses 2 nd EPS/Metop RAO Workshop 20-22 May 2009 Barcelona, Spain page 1
GOALS Surface processes modeling Better understanding of the West African Monsoon and its interactions with global environment Impact of climatic variability on surface processes Role of continental surfaces on climate Radar sensors to assess: Vegetation biomass Net Primary Productivity (NPP) Ë Carbon budget Soil moisture in relation with: evaporative fluxes moisture deeper layer To assess the seasonal and inter-annual variations of these 2 key surface parameters over the Sahel from spaceborne radars sensors: scatterometers. 2 nd EPS/Metop RAO Workshop 20-22 May 2009 Barcelona, Spain page 2
The SAHEL Alternation of a long dry season (October - June) and a short rainy season (June September) Vegetation closely linked to the rainfall regime 2 nd EPS/Metop RAO Workshop 20-22 May 2009 Barcelona, Spain page 3
Sahelian Pastoral Area herbaceous steppe (annual grass) : herbaceous layer (0-50%) shrub layer (0-20%) trees (1-5%) 2 nd EPS/Metop RAO Workshop 20-22 May 2009 Barcelona, Spain page 4
Strong interannual variability 1984 1985 1986 1987 1988 1989 1990 1991 1999 2 nd EPS/Metop RAO Workshop 20-22 May 2009 Barcelona, Spain page 5
Radar signal temporal evolution over a Sahelian region (ERS scatterometer Rharous site) Signal stability during the ERS mission Annual cycle linked to the alternation rainy/dry season Inter-annual variability Incidence angle: 45 2 nd EPS/Metop RAO Workshop 20-22 May 2009 Barcelona, Spain page 6
Study area : GOURMA (MALI) Rharous Gossi Hombori Seno Pastoral areas 2 nd EPS/Metop RAO Workshop 20-22 May 2009 Barcelona, Spain page 7
ERS scatterometer Temporal evolution of s 0 (45 ) : 4 sites v Signal inter-annual variability v Signal level during dry season southward southward 2 nd EPS/Metop RAO Workshop 20-22 May 2009 Barcelona, Spain page 8
Calibration with in situ biomass measurements Modeling study Sahelian vegetation Growth model (STEP) Daily surface parameters : Biomass Soil moisture Radiative Transfer model (Karam, 1992) Simulated Backscattering coefficient s 0 (t) Daily meteorological data Soil parameters Soil roughness Vegetation parameters h c 2 nd EPS/Metop RAO Workshop 20-22 May 2009 Barcelona, Spain page 9
Simulation and ERS scatterometer observations Ë Overall good concordance Incidence angle: 45 2 nd EPS/Metop RAO Workshop 20-22 May 2009 Barcelona, Spain page 10
SURFACE PARAMETER ESTIMATION METEOSAT Rainfall estimation Sahelian vegetation growth model (STEP) Daily surface parameters : Biomass Soil moisture Radiative Transfer model (Karam, 1992) s 0 (t) simulations OPTIMISATION Observations : Satellite measurements (s 0 ) 2 nd EPS/Metop RAO Workshop 20-22 May 2009 Barcelona, Spain page 11
Inversion method : Biomass production map at a regional scale Biomass production map: 2000 rainfall estimation: METEOSAT Biomass + soil moisture: ËERS WSC + METEOSAT Seno Validated over 4 pastoral sites (Gourma and Seno regions) Absolute errors: ~ 200 kg Dry Matter (DM) / ha T Relative error : 17 % (kg Dry Matter / ha) 2 nd EPS/Metop RAO Workshop 20-22 May 2009 Barcelona, Spain page 12
Spatialisation to the Sahel Maximum herbaceous mass, year 1999 (0 3.5 t DM/ha) SNDVI VEGETATION ERS WSC Overall good agreement between both sensors Detection of low biomass values Wrong values over swampy areas 2 nd EPS/Metop RAO Workshop 20-22 May 2009 Barcelona, Spain page 13
Spatialisation to the Sahel - Validation sites Agro-Pastoral Area: Fakara (NIGER) ERS WSC Pastoral Areas: Ferlo (SENEGAL) Gourma (MALI) Ë In situ measurement (rainfall, biomass) available 2 nd EPS/Metop RAO Workshop 20-22 May 2009 Barcelona, Spain page 14
TEMPORAL EVOLUTION of ERS SCATTEROMETER RESPONSE ( Sahelian sites) Gourma Rharous (Mali) Ferlo (Senegal) 2 nd EPS/Metop RAO Workshop 20-22 May 2009 Barcelona, Spain page 15
COMPARISON BETWEEN ERS SCATTEROMETER RESPONSE AND SIMULATIONS FERLO (SENEGAL) Good overall concordance Ë Reliability of spatialisation of surface parameters estimation over the Sahel 2 nd EPS/Metop RAO Workshop 20-22 May 2009 Barcelona, Spain page 16
QUIKSCAT (Ku band) TEMPORAL SIGNATURES ËAtmospheric perturbation hampering interpretation +: s 0 measured (L2A) *: s 0 atmospheric attenuation corrected (L2A) x: BYU products (Brigham Young University) 2 nd EPS/Metop RAO Workshop 20-22 May 2009 Barcelona, Spain page 17
Long time series provided with METOP - ASCAT measurements FERLO ( SENEGAL) *: ERS scatterometer *: METOP - ASCAT 2 nd EPS/Metop RAO Workshop 20-22 May 2009 Barcelona, Spain page 18
Long time series provided with ASCAT mission FERLO ( SENEGAL) *: ERS scatterometer *: METOP - ASCAT ASCAT acquisition: Higher temporal frequency 2 nd EPS/Metop RAO Workshop 20-22 May 2009 Barcelona, Spain page 19
ASCAT DATA - Comparison between 25 km and 50 km spatial resolution products FERLO ( SENEGAL) Ds 0 < 0.2 db Ë 25 km products for land surfaces monitoring 2 nd EPS/Metop RAO Workshop 20-22 May 2009 Barcelona, Spain page 20
CONCLUSIONS The method is relevant for a spatialisation over the whole Sahel The method requires Meteosat data and ASCAT data for the surfaces C band leads to better results than Ku band ASCAT data warranty the follow on of ERS data ASCAT data dispersion is lower ASCAT data with a better temporal resolution 25 km products are relevant 2 nd EPS/Metop RAO Workshop 20-22 May 2009 Barcelona, Spain page 21